InetSoft BI Webcast: Data Mashup and Geographic Mapping

Below is more from Information Management’s Webcast by DM Radio, “The Last Mile: Data Visualization in a Mashed-Up”. This Webcast was hosted by Eric Kavanagh and included BI consultants William Laurent and Malcolm Chisholm, and InetSoft's Product Manager Tibby Xu.

EK: Malcolm, let’s bring you in here as well. What do you think of our ideal scenario?

Malcolm Chisholm (MC): Well, I think I would go a little further actually. One of the things I have been wrestling with is taking what Bill said about GIS, being the most common application where geography is the fixed frame of reference. But even in financial services or even doing data administration, data governance, for instance, we need to go further so we need to do something about developing a fundamental canvas for a mashup. For instance, I’ve been doing some data discovery work recently, and I’d love to be able to set that into some kind of mashup, but I need something that visualizes the production data landscape. Maybe it could be by subject area, I don’t know.

I think you have to move beyond GIS. The thing about geographic mapping is that we’re all familiar with it. We all know what it is. There isn’t a semantic challenge to understanding it. But let’s say if I was to create the data topography of the production landscape. Then I am going to have to create some kind of canvas and communicate to people in a fairly clear way what that is. Subject area is probably a good way to go. Then I can start to go overlay things on it like applications, servers, databases, flows of data, whatever. I think a big challenge in that, however, I think that is probably something that is going to be met. I think that there is tremendous demand for it.

EK: Tibby, can you tell us about a case study of a company using data mashup and geographic mapping?

In the competitive landscape of ride-sharing services, companies are constantly seeking innovative ways to enhance operational efficiency, optimize routes, and improve the overall customer experience. One leading ride-sharing service (we will refer to as "RideShare Co.") faced challenges typical to the industry, such as managing vast amounts of real-time data, ensuring accurate and efficient route planning, and providing timely services in congested urban areas. To address these challenges, RideShare Co. turned to data mashup techniques combined with geographic mapping, enabling them to integrate multiple data sources and visualize insights in a manner that significantly improved decision-making processes and operational outcomes.

Background

RideShare Co. operates in several major cities around the world, offering a platform that connects passengers with drivers via a mobile app. The company's operations are heavily reliant on data—ranging from real-time traffic information and weather conditions to customer demand patterns and driver availability. However, managing and making sense of this data in real-time posed significant challenges. The data was scattered across different systems and formats, making it difficult to gain a comprehensive view of the operational landscape. Moreover, the dynamic nature of urban traffic, coupled with unpredictable customer behavior, required a solution that could not only aggregate but also analyze and visualize data on the fly.

The Challenge

The primary challenges faced by RideShare Co. included:

1. Data Integration: The company needed to combine data from various sources, such as GPS data, traffic reports, weather forecasts, customer feedback, and driver availability, into a single coherent platform. This data was often stored in disparate systems, making it difficult to perform real-time analysis.

2. Real-Time Route Optimization: Efficient routing is critical in the ride-sharing industry, where minimizing wait times and travel durations can significantly impact customer satisfaction and operational costs. However, the dynamic nature of urban traffic, coupled with changing weather conditions, made it challenging to optimize routes in real time.

3. Demand Prediction: Understanding and predicting customer demand in different locations at different times was essential for deploying drivers effectively. However, the variability in demand due to factors like weather, events, and even socio-economic conditions made it difficult to predict with accuracy.

4. Enhanced Customer Experience: Providing accurate ETAs, reducing wait times, and offering personalized experiences were key to retaining customers in a highly competitive market. Achieving these goals required a deep understanding of customer behavior and preferences, which in turn relied on effective data analysis and visualization.

Solution: Data Mashup and Geographic Mapping

To overcome these challenges, RideShare Co. implemented a data mashup solution that integrated data from multiple sources into a unified platform. This platform was enhanced with geographic mapping capabilities, enabling the company to visualize data in a spatial context, which is crucial for operations based on geographic location.

1. Data Integration and Mashup:
   • RideShare Co. deployed a data mashup tool that allowed them to pull in data from various internal and external sources. Internally, the platform integrated GPS data from vehicles, customer booking data, driver availability, and historical ride information. Externally, it pulled in real-time traffic updates, weather forecasts, event schedules (such as concerts or sports games), and even public transport data.
   • This mashup of data sources provided a holistic view of the operational environment. For instance, by combining weather data with historical ride patterns, the company could predict increased demand during rainstorms and proactively position drivers in areas likely to experience higher ride requests.
2. Geographic Mapping:
   • The integration of geographic mapping tools was central to the solution. Using advanced geographic information system (GIS) technology, RideShare Co. was able to visualize data in a spatial format, overlaying different data layers such as traffic congestion, driver locations, and customer demand hotspots onto city maps.
   • This spatial visualization allowed dispatchers and the algorithm-based system to make more informed decisions about where to position drivers, how to route them efficiently, and where to anticipate spikes in demand. For example, during a large concert, the system could identify nearby drivers and suggest optimal routes that avoided traffic jams, ensuring that they arrived at pick-up points quickly.
3. Real-Time Route Optimization:
   • The geographic mapping system was integrated with real-time traffic and weather data, enabling the ride-sharing platform to dynamically adjust routes based on current conditions. If a sudden traffic jam or road closure occurred, the system could instantly reroute drivers, minimizing delays.
   • This capability was particularly beneficial during peak hours when traffic congestion is highest. By continuously monitoring and adjusting routes, RideShare Co. was able to reduce average trip times and improve the accuracy of ETA predictions, enhancing customer satisfaction.
4. Demand Prediction and Resource Allocation:
   • The data mashup platform also supported advanced analytics for demand prediction. By analyzing historical data alongside real-time inputs (such as weather or special events), the system could forecast where and when demand for rides would peak.
   • These predictions were visualized on the geographic maps as heatmaps, showing areas of expected high demand. This allowed RideShare Co. to strategically deploy drivers to these areas in advance, reducing wait times for customers and increasing the number of completed rides per driver.
5. Enhanced Customer Experience:
   • By integrating data mashups and geographic mapping into their operations, RideShare Co. significantly improved the customer experience. The accurate ETAs, shorter wait times, and smoother rides (thanks to optimized routing) led to higher customer satisfaction ratings.
   • Additionally, the platform's ability to personalize service based on data insights (such as suggesting preferred routes or offering promotions based on past behavior) helped RideShare Co. differentiate itself in a crowded market, driving customer loyalty and retention.

Previous: Data Mashups vs. Data Warehouses

Next: How Data Mashups Work